The overarching goal of this project is to understand two important aspects of epigenetic control during medulloblastoma formation, namely the difference between cancer stem cells and bulk tumor, and the role of DNA methylation dependent gene silencing in tumor progression. Our experimental rationale that connects these two questions is an observation we made that normal mouse embryonic stem cells do not require the sole maintenance DNA methyltransferase Dnmt1, whereas all known somatic cells tested require this gene for survival. The increasing awareness that pluripotency genes, those genes important for embryonic stem cell self-renewal, may play an important functional role in cancer stem cells encourages us to test the importance of this pathway in brain tumor stem cells derived from medulloblastoma. To approach this question, we are utilizing a mouse model harboring mutations in two tumor suppressor genes, Ptch and p53, that when mutated lead to a highly penetrant and lethal medulloblastoma. To examine the role for epigenetic control in cancer stem cells we will first characterize Ptch-P53 medulloblastomas to determine whether they contain a self- renewing cancer stem cell population. We will then test whether the self-renewing cells within these medulloblastomas are Dnmt1 dependent like other somatic cells or are Dnmt1 independent like embryonic stem cells. To address the role of epigenetic silencing, we will utilize a viable hypomorphic allele of Dnmt1 and ask whether genomic DNA hypomethylation induces or suppresses medulloblastoma formation. Depending on tumor context, Dnmt1 hypomorphic mice have revealed important roles for DNA hypomethylation in promoting chromosomal instability and alternatively in preventing aberrant gene silencing. We will determine whether chromosomal or epigenetic mechanisms prevail in medulloblastoma progression using a genetic approach with Dnmt1 hypomorphic mice. Finally, we have recently demonstrated that transient demethylation in embryonic stem cells leads to global loss of imprinting, and loss of p53 expression in derived mouse fibroblasts through an unknown mechanism. We have derived ES cells from medulloblastoma prone Ptch mice in order to carry out transient demethylation in stem cells and test whether epigenetic mechanisms lead to loss of p53 and promote medulloblastoma tumor progression in vivo. These results will extend prior results that global loss of imprinting can promote tumorigenesis, and provide a model to dissect the mechanism through which transient epigenetic changes exert long term effects on cell growth regulation, and tumor predisposition in vivo. PUBLIC HEALTH RELEVANCE: This proposal tests four hypotheses regarding the contribution of epigenetic alterations to medulloblastoma initiation and progression. We ask are cancer stem cells similar to embryonic stem cells in their lack of dependence on the DNA methyltransferase Dnmt1, and does global transient demethylation promote cancer by potentiating cancer stem cell expansion? We test the contribution of epigenetic alterations by asking does gene silencing contribute to tumor initiation or progression, and are imprinted genes a key functional class of epigenetic targets that become deregulated in these pediatric brain tumors? Answers to these questions should yield insight into the origins of the most common pediatric brain tumor, and will likely impact a broader understanding of cancer epigenetics.